Re: Where do I begin?

From: Steve Werby <steve-lists_at_befriend.com>
Date: Thu 22 Aug 2002 23:03:23 -0500

"John Smart" <JSmart_at_InternetDesign.com> wrote:
> Either I missed the start of this, or I am being very stupid (let's be
> honest here, a combination is likely!) but I am unsure of the goal.

Lyn's goal was to improve the position of her database driven site's pages
in search engines. Comments were made that some search engines do not index
pages with query strings in the URLs and those that do may not rank them as
high as like pages without query strings.

> You either want to make dynamic pages look like static pages

I assume you mean URLs, not pages. I agree there are advantages to doing
so. Whether the content on a page looks static or dynamic to the human eye
is a different topic entirely, but I assume that's not what you meant.

>, or you want to
> take dynamic pages and convert them to static pages. I see no real reason
> for doing the latter - the dynamic content will be lost, so the static
> content created will presumably go out of date.

Not so. In fact, I have built several sites where dynamic content is
regularly converted into static pages. Why? Because the page content for
the dynamically generated pages either doesn't change or changes
infrequently and I could reduce the load on the server while serving up
pages much more quickly. Of course, the benefits will be more pronounced on
a high traffic site and/or a site with complex dynamic pages that use a lot
of CPU cycles and take longer than desired to load. On some sites I just
run a regularly scheduled script periodically to loop through all pertinent
database records and rebuild all pages. On other sites, I loop through the
database records, checking for flags or timestamps and only update pages
that have changed since the previous run of the script. For example, I had
a client who ran an affiliate hotel directory site. The site had tens of
thousands of dynamic pages that were generated by a Perl script that pulled
data from a database, formatted it and displayed it. With 200,000 page
views per day on a low end server, the load was significant during peak
times. But the content on each page only changed when a new dataset was
loaded once per month. So I wrote a script to loop through the database
after the dataset was loaded and generate a static HTML page for each hotel.
Instead of running the script 6 million times per month as the user loaded
the page, my script was run once per month. As a result the time required
for the webserver to generate a page during peak load went from about 6
seconds to 0.2 seconds.

> It is possible to "hide" dynamic content - all of my site
> (www.internetdesign.com) is dynamic, but 90%+ of the pages end .html - I
> just told Apache to treat .html as .php.

Dynamic content per se is not the issue. The issue is that URLs with query
strings will be ignored by some SE spiders and will result in lower SE
positioning in some SEs. IMO, simply changing extensions from .php, .jsp,
.pl, .cgi, .asp or something else associated with a scripting language to
.htm or .html will accomplish little. It's whether the URL contains a query
string and .htm and .html files can contain them too.

> This could cause a server slowdown
> if php was trying to parse all these pages and there was no need to, but
> given that all the pages have php content there is no loss of quality.

What you say is true. In reality, on servers with little traffic or CPU
cycles being used for other processed, the speed difference and CPU load
difference between an HTML page parsed as PHP and one that's not parsed is
negligible. And in high load situations, it's easy to overcome this by
configuring the webserver to only parse specific .html files or files in
specific directories or sites as PHP.

> Feel
> free to mail me for details of this - I don't want to bore the entire
> subscriber base with my nerdy ramblings!

Isn't the point of this mailing list to share ideas and knowledge?

> To convert a site into static pages all you need to do is grab the site,
and
> possibly rename all the files (assuming that you are grabbing .asp (for
> example) and want .html. Changing every link and every file name should
not
> be too hard - using mass search and replace tools (textpad would do the
job)
> and a file renaming program (there are many available).

True, this isn't a difficult process, but it wouldn't be effective for any
sites with query strings. So it definitely wouldn't help with Lyn's
(original poster) site at http://www.agiftgallery.com/. Simply changing
host.domain.tld/page.asp?var1=value1&var2=value2 to
host.domain.tld/page.html?var1=value1&var2=value2 will do little to nothing
to address the problem Lyn is facing. And changing it to
host.domain.tld/page.html will effectively break the site.

So there are benefits of converting dynamic pages to static pages and there
are benefits to making site and webserver changes to allow dynamic pages to
have URLs that appear to be static. But whether you use one over the other
or both together really depends on what you are trying to accomplish. I
hope this was informative and clear. Feel free to ask for clarification if
necessary.

--
Steve Werby
President, Befriend Internet Services LLC
http://www.befriend.com/






Received on Thu Aug 22 2002 - 23:03:23 CDT